Spiny neurons throughout the nervous system have thornlike protuberances arising from their dendrites. These dendritic spines are a site of asymmetric excitatory inputs to the neurons. Spine shape is thought to affect the efficacy of the synaptic input. The focus of this proposal will be to study the ultrastructure of dendritic spines on three different types of spiny neurons. The major emphasis will be on the pyramidal cells of area CA1 in the rat hippocampus. Comparison measurements will be made on granule cells of area dentata of the same hippocampus, and on cerebellar Purkinje cells of the same brain. I will use computer-assisted methods to reconstruct dendritic spines from serial electron micrographs. Then measurements of spine length, neck diameter, surface area and volume will be made. Synaptic surface area, the volume of spine cisternae and the diameter of the parent dendrite will also be measured. Then these measures will be used in computer programs that model how relationships among these structural features might affect synaptic efficacy. Ultrastructural studies of dendritic spines suggest that their necks shorten and widen following high frequency afferent activation. This structural change should increase the efficacy of synapses terminating on the spine by decreasing the resistance to synaptic current flow through the spine neck to the dendrite. In the hippocampus, high frequency stimulation of afferent input results in a long-lasting incrase in the physiological response of the postsynaptic cell. This change in response is referred to as long-term potentiation (LTP) and represents a candidate mechanism for information storage. In area CA1 of the rat hippocampus, the magnitude of LTP produced is about three times greater at postnatal day 15 than in the adult (Harris and Teyler, In Press). Spines will be studied to determine whether changes in there ultrastructure during development would account for this age related difference in LTP magnitude. In addition, LTP will be induced in the apical dendritic field of area CA1 in hippocampal slices at days 15 and 60. Spines in the vicinity of stimulation will be reconstructed and measured. Comparison measurements will be made on spines in the basillar dendritic field of the same slice, and on apical spines of non-potentiated slices. In order to unambiguously identify spines that contact stimulated axons, these axons will be filled with horse-radish peroxidase following the induction of LTP.